Bradyarrhythmia secondary to vagus nerve stimulator 7 years after placement

Abstract

We present a case of a 38-year-old man with a previous medical history of asthma and refractory epilepsy requiring vagal nerve stimulator (VNS) placement 7 years prior to the presentation who was found to be in atrial fibrillation with a rapid ventricular response during a preoperative evaluation, which prompted transoesophageal echocardiography and subsequent cardioversion. In preparation for cardioversion, the VNS was turned off and the patient was cardioverted to normal sinus rhythm. Following cardioversion, the VNS was activated again. During recovery, the patient was experiencing several episodes of first-degree and second-degree Mobitz type-II atrioventricular (AV) block. In response, the VNS was deactivated indefinitely. On interrogation of a loop recorder 2 weeks after discharge, the patient did not have any further evidence of AV conduction delay.

Background

Vagal nerve stimulation therapy is a non-pharmacologic treatment alternative for patients with refractory epilepsy. The Vagal Nerve Stimulator (VNS) device is a battery-powered generator that is surgically implanted to deliver electrical impulses to the left vagus nerve in the carotid sheath via stimulating leads. Since the right vagus nerve innervates the sinoatrial node (SA node), electrical stimulation of the left vagus nerve is preferred in order to avoid adverse cardiac conduction effects. Nevertheless, patients with cardiac conduction disorders at their baseline are considered poor candidates for vagal nerve stimulation therapy. Although the mechanism of seizure reduction is not well understood, vagal nerve stimulation therapy has demonstrated efficacy in patients with a broad range of seizure types. The most common side effects of vagal nerve stimulation therapy are hoarseness, throat pain, coughing, shortness of breath, tingling and muscle pain. In 0.1% of cases, patients experience bradycardia during initial intraoperative lead testing following VNS device implantation. In addition, a recent case series described three cases of transient ventricular asystole with a complete heart block during the intraoperative lead testing. There have also been four recently published case reports of late-onset VNS-induced episodes of bradycardia and atrioventricular (AV) block in patients who presented with dizziness, unsteadiness and syncope. The timeframe between VNS device implantation and the presenting symptoms of AV block ranged from 1 to 12 years. Unfortunately, the symptoms of an AV block can be mistaken for new-onset seizure types. The consequences of an undiagnosed AV block without a junctional escape rhythm can be sudden cardiac arrest and death. Therefore, a late-onset AV block is an extremely rare complication of vagal nerve stimulation therapy that needs to be identified quickly in patients with refractory epilepsy who present with syncopal episodes.

Case presentation

We present a case of a 38-year-old man with a previous medical history of asthma and epilepsy requiring VNS placement 7 years prior to admission. Since stopping antiepileptic therapy 3 years prior to admission, the patient had been seizure-free. Several months prior to admission, the patient was found to have bradycardia, which prompted loop recorder implantation and further evaluation. Loop recorder interrogation done later in the hospital course did not show any arrhythmia prior to the presentation to our hospital.

On the day of admission, the patient was found to have evidence of atrial fibrillation with a rapid ventricular response (150–160 bpm) (figure 1) while undergoing a preoperative evaluation for the drainage of a dental abscess. Although he was asymptomatic, he was advised to go to the emergency department for further evaluation. In the emergency room, he received 25 mg of diltiazem via intravenous push and was started on an intravenous diltiazem infusion at a rate of 10 mg/hour. Complete blood count, basic metabolic panel, thyroid function studies and troponin-I level revealed no acute abnormalities.

Figure 1.

ECG done at the time of admission with atrial fibrillation with rapid ventricular response.

In the emergency department, the on-call cardiologist evaluated the patient. At that moment in time, the patient had evidence of atrial fibrillation with a heart rate of 110 bpm. He was started on anticoagulation therapy and the intravenous diltiazem infusion was increased to 15 mg/hour.

On admission to the medical floor, the patient experienced several episodes of sinus arrest. The first episode was 2.2 s in duration and was accompanied by a heart rate of 40–45 bpm. The intravenous diltiazem infusion was decreased to 10 mg/hour and then to 7.5 mg/hour. The second episode of sinoatrial arrest was 2.0 s in duration. The intravenous diltiazem infusion was then decreased to 5 mg/hour. Several hours later, the patient experienced two episodes of sinoatrial arrest (3.3 s and 4.4 s in duration, respectively). The intravenous diltiazem infusion was stopped.

First episode of Mobitz type-II AV block was noticed about 17 hours after intravenous diltiazem drip was stopped. Neurology team was consulted who provided clearance to deactivate the VNS in order to proceed with transoesophageal echocardiography (TEE) and cardioversion. Immediately after the VNS was turned off by the manufacturing company, the patient’s systolic blood pressure increased to 180–190 mm Hg (on several confirmed readings). In response, the VNS was turned on and the patient’s blood pressure was normalised. The TEE was then performed while the VNS was active. In preparation for cardioversion, the VNS was turned off again. We avoided the use of propofol for anaesthesia to avoid potential bradycardia as patient was having bradycardia and sinus arrest while on diltiazem drip, and only midazolam was used. The patient underwent cardioversion with 150 J of biphasic current and returned to normal sinus rhythm (figure 2).

Figure 2.

ECG with normal sinus rhythm following electrical cardioversion.

Following cardioversion, the VNS was activated again, and the patient was found to have sinus bradycardia with heart rate of 40–50 bpm. The decision was made to monitor the patient overnight on the telemetry service. During this period of time, the patient experienced multiple episodes of first-degree and second-degree Mobitz type-II AV block (figure 3). Electrical cardioversion as a potential cause of AV blocks was ruled out as patient was noted to have first episode of AV block before electrical cardioversion was performed. In an effort to differentiate between the potential causes of advanced AV block in this patient (intrinsic cardiac vs VNS related), the VNS was turned off indefinitely following neurology consultation. Possible side effects of remaining off VNS, including aggravation of recurrent seizures, were discussed with the patient. The patient was also started on 1 g of levetiracetam every 12 hours for seizure prevention.

Figure 3.

Telemetry recording with Mobitz type-II heart block noted overnight following cardioversion.

The patient was monitored for 1 day. Since his vital signs remained stable, he was discharged home with instructions to follow-up with his cardiologist and neurologist.

Investigations

Transthoracic echocardiogram

Left ventricle: normal size, thickness and systolic function. Ejection fraction (EF) 60%–65%. Normal filling pattern.

Left atrium: normal size. Atrial appendage velocities are normal. No thrombus. Interatrial septum is intact.

Right atrium: normal size.

Right ventricle: normal size.

Aortic valve: normal structure. No stenosis or regurgitation.

Mitral valve: grossly normal. No stenosis or regurgitation.

Tricuspid valve: not well visualised but grossly normal. Trace to mild regurgitation.

Pulmonic valve: no regurgitation.

Arteries: aortic root normal size.

Venous: pulmonary flow patterns are normal. Inferior vena cava (IVC) normal size and collapsibility.

Pericardium/pleura: no effusion.

Differential diagnosis

Our main two differential diagnoses were advanced block secondary to intrinsic cardiac cause/extrinsic cause from VNS versus advanced AV block secondary to use of diltiazem.

At the time of admission to the hospital, our patient was started on intravenous diltiazem for initial rate control, but due to the development of several episodes of bradycardia and sinus arrest, diltiazem was stopped. There was a significant gap of 17 hours between stopping intravenous diltiazem and electrical cardioversion following which advanced AV block was observed. As duration of action for intravenous diltiazem is 0.5–10 hours, we ruled out the use of diltiazem as potential cause of AV blocks. No other AV nodal blockers were used including amiodarone in our patient’s case.

Treatment

1. Diltiazem for rate control.

2. Electrical cardioversion for return to normal sinus rhythm.

3. Deactivation of the VNS indefinitely.

4. Administration of 1 g of levetiracetam every 12 hours for seizure prevention.

5. Anticoagulation therapy with rivaroxaban for 4 weeks.

Outcome and follow-up

Following deactivation of the VNS, the patient returned to normal sinus rhythm and demonstrated no evidence of bradycardia including AV block. Two weeks after discharge, the patient followed up with his cardiologist at the outpatient clinic. Loop recorder interrogation at that time did not show any evidence of episodes of bradycardia or sinoatrial arrest.

Discussion

Approximately 1% of the global population has some form of epilepsy, with partial seizures being the most common in adults.¹ VNS is a therapeutic alternative for cases of epilepsy that have become refractory to antiepileptic therapy or patients who cannot tolerate the side effects of antiepileptic agents.²

The mechanism of action of vagal nerve stimulation therapy was initially believed to be related to an increase in the extracellular concentration of norepinephrine. However, it is now known that vagal nerve stimulation therapy increases the concentration of gamma-aminobutyric acid (GABA) in the cerebrospinal fluid.¹ While vagal nerve stimulation therapy is an effective treatment option for epilepsy, uncommon, mild, transient complications can arise during the initial VNS and perioperative period.^{2 3}

The most common complications are hoarse voice, throat pain, cough, dysphonia (vocal cord dysfunction), paraesthesia, superficial skin infections, neck pain, dysphagia, dyspnoea and headache.^4–6 Cardiac arrhythmias are uncommon complications which tend to present in the perioperative period, especially during the initial lead testing with 1 mA for 500 µs and 20 Hz for 60 s.² The most common cardiac abnormalities noted during this period are complete heart block, sinus bradycardia and ventricular asystole.^{2 3}

Adverse cardiac effects, especially bradycardia secondary to AV block, are more common during lead testing, but there are documented cases of arrhythmias that appear years after VNS placement.⁷ Following VNS placement, the timeframe for the appearance of adverse cardiac effects appears to range from 2 to 9 years.⁴

Some studies have also compared and documented the changes in blood pressure during the Valsalva manoeuvre and at rest in patients undergoing vagal nerve stimulation therapy. These patients have demonstrated an increased systolic blood pressure during phases IIb and V of the Valsalva manoeuvre and decreased blood pressure at rest 1 month after VNS placement.⁷ This phenomenon might explain our patient’s episode of hypertension following the deactivation of his VNS for cardioversion. In addition, another case report presented a patient from Houston with episodes of AV block during the night and episodes of atrial fibrillation during the day.⁸ These complications resolved without reoccurrence after the VNS was deactivated.⁴

But why do adverse cardiac events arise in patients with VNSs? The right vagal nerve has a greater negative chronotropic effect on the sinoatrial node, while the left vagal nerve has a similar effect on the AV node. Therefore, direct activation of the parasympathetic pathway can exaggerate the effects on the AV node.²

In addition, the VNS activates the ipsilateral and contralateral solitary nuclei, perhaps by direct vagal stimulation. The solitary nucleus receives innervation from multiple areas such as the carotid sinus nerve, aortic depressor nerve, cranial nerves V, VII and IX, the area postrema (indirectly affects the cardiovascular reflexes), the rostral ventrolateral medulla, the parabrachial nucleus of the pons, the amygdala, periventricular nucleus and the lateral and posterior hypothalamus.²

The solitary nucleus projects to the parabrachial nucleus, reticular activating system, locus ceruleus and periaqueductal grey matter. It is therefore involved in modulating pain and respiratory pathways in conjunction with the cardiovascular reflexes.²

Furthermore, stimulation of the vagal efferent pathway that innervates the heart leads to decreased pacemaker activity, AV nodal conduction and the excitability of the His-Purkinje system. However, this does not explain why adverse cardiac events appear in only a subset of patients receiving vagal nerve stimulation therapy. It may be that there are considerable anatomic variations in the vagal innervation of the SA and AV nodes such that inadvertent electrode placement can stimulate more than just the A and B vagal cardiac nerves and lead to direct vagal stimulation of the heart.3–5 8–10

It has been hypothesised that collateral current spread can also stimulate cervical branches of the vagus nerve.⁷ It may be that years after implantation, the sensitivities of the cardiac mechanoreceptors or the solitary nuclei are upregulated and lead to late presentation of adverse cardiac effects, as seen in our patient.^{6 11} Factors that are postulated to be involved in the modulation of cardiac mechanoreceptor sensitivities are antiepileptic medications, anaesthesia and personal comorbidities (diabetes mellitus, hypertension, etc).² Finally, human error, such as lead polarity reversal and technical malfunction of the VNS device, can lead to adverse cardiac events.²

It has been proposed in the literature that institutions offering vagal nerve stimulation therapy for epilepsy should follow a protocol that involves performing a lower intensity testing of the leads if symptoms appear during placement. It has been suggested to challenge the patient with 0.25 mA for 30 s. If the patient tolerates the initial test, proceed to challenge them with 1 mA for 90 s. If the patient tolerates this testing, traditional testing may be performed.² Following this protocol may lead to a decreased incidence of long-term complications.

In conclusion, the adverse cardiac effects of vagal nerve stimulation therapy can be lethal. Therefore, providers caring for patients with VNSs should be aware of and remain vigilant for the complications associated with this therapeutic alternative. This notion even applies to patients who have tolerated their vagal nerve stimulators and remained asymptomatic for many years.

Learning points.

• Vagal nerve stimulation therapy is a therapeutic alternative for refractory epilepsy that is generally well tolerated but is associated with fatal arrhythmias in rare circumstances.

• Deactivation of the vagal nerve stimulator is diagnostic and therapeutic in patients experiencing adverse cardiac effects.

• The pathophysiology of adverse cardiac effects in the setting of vagal nerve stimulation therapy seems to be related to activation of the parasympathetic efferent fibres that innervate the heart, device malfunction/displacement and native or acquired anatomic variations.